Such a device is disclosed in WO95/11755 for "sequencing by hybridisation" applications. The device comprises a substrate provided with channels, the channels being oriented substantially perpendicular to the surface of the substrate. Three types of substrate are disclosed. The first type is comprised of a multitude of hollow glass fibres. It is manufactured by stacking glass fibres having an etchable core, providing the stack with flat ends, polishing those ends, and etching the cores, usually with acid. The second type of substrate is produced by electrochemical etching of a crystalline silicon wafer. First, the position of the channels as well as their size are defined using standard photolithographic methods. Subsequently the oriented channels are formed electrochemically. The third type of substrate is produced by nuclear track etching of an inorganic substrate. This method, comprising the steps of exposing the substrate to heavy, energetic charged particles and wet-etching, results in a substrate with channels scattered randomly over the surface of the substrate. With higher pore densities and porosity there is more chance of fusion of channels, which show reduced flow resistance with respect to other, non-fused channels.
All three types of substrates are quite expensive because of the labour-intensive manufacturing processes and/or expensive starting materials and wasteful operations, such as sawing and polishing, and/or expensive equipment. In addition, the substrates are characterised by a relatively low porosity of 30% and more. More advantageous, higher porosities of up to 80% are said to be achievable, but only at relatively low channel densities, with the disadvantage that the effective surface area of the channels of a particular area of the substrate is lower in comparison with a substrate having a comparable porosity but with higher channel densities (and consequently narrower channels). A further disadvantage of the silicon-based substrates as disclosed in WO 95/11755 is that they are not transparent for light. These substrates therefore prohibit the advantageous use of optical marker systems for the detection of analyte bound in the substrate. Popular optical marker systems are for instance based on enzymatically induced colour reactions, on bio- or chemi-luminescence, or on photoluminescence. In the latter case both the excitation light and emitted luminescent light have to pass through the substrate material.
The object of the present invention is to overcome the above disadvantages and provide a substrate having both a high channel density and a high porosity, allowing even higher density arrays comprising different first binding substances per unit of the surface for sample application. In addition, the substrate is highly transparent for visible light. More in particular, the object of the present invention is to provide a device comprising a relatively cheap substrate that does not require the use of any typical microfabrication technology and, that offers an improved control over the liquid distribution over the surface of the substrate.
The above objects are achieved with a device wherein the porous substrate is an electrochemically manufactured metal oxide membrane.
Metal oxide membranes having through-going, oriented channels can be manufactured cheaply through electrochemical etching of a metal sheet. Metals considered are, among others, tantalum, titanium, and aluminium, as well as alloys of two or more metals and doped metals and alloys. The metal oxide membranes are transparent, especially if wet, which allows for assays using various optical techniques. Such membranes have oriented channels with well controlled diameter and advantageous chemical surface properties.